Let's try it this way.
Proposition:
The only way to scale a blockchain is to add an additional layer that allows off-chain transactions.
Explain to me in less than 500 words why that statement is false. I'm not interested in what you think is wrong with LN but why another layer isn't the solution. That is the topic of this thread.
Answer:
http://satoshi.nakamotoinstitute.org/emails/cryptography/2/I didn't count Satoshi's words, I hope it is less than 500.
Satoshi proposed a client/multi-server network ; we observe that bitcoin BECAME a client/multi-server network. A client/multi-server network can perfectly scale, as Satoshi explained above.
BTW, I'm not AGAINST the LN network, and I can clearly say that Carliton Banks has induced me to clear up a misunderstanding I had about it, which I recognized without problems, and makes me see the LN as way more useful than I thought. But I think that the LN is not a payment network, and that payment is not a good application (like payment is not a good application for bitcoin either).
The LN network is a good thing to make a fast high-frequency multi-asset speculative market. I'm quite bullish on that. I think that if it takes off (and I think it will), it can be much more destructive than classical speculative finance has ever been (even making the 1929 crisis, and its consequence, WWII, look like small beer). I'm really keen on that idea.